I wasn't really saying they could succeed, I was more of referring to a company having a competitive advantage in a specific area, with that area growing they could benefit.
Personally, a dual-function CPU doesn't seem to make sense for IoT. I think it'll be ARMy vs. Intel/AMD in that segment, like the ULP market above.
VIA may well play ARM products in IoT, but they'll run up against a million other companies doing the same. Not sure what they'll bring to the table here.
I think there's very little actual product overlap between VIA ARM SoCs, which are made by WonderMedia (and use very low end and old apps processors like ARM11), and their x86 CPUs which are made by Centaur. VIA doesn't really seem that great at integrating its subsidiaries in general, which could explain why Nano CPUs have yet to end up on the same die as Chrome GPUs.
In terms of patents, I should have clarified only a varying percentage are used, especially in end products. But even then, some patents can contribute indirectly (one may have led to another).
Sure. Patents can give a hint at what might be released. And they definitely give a hint at what the company is or at least was working on. But too many people are using them as a crystal ball to tell the future with perfect accuracy.
In this case it looks like VIA legitimately put some effort into figuring out how to make their Nano core work with ARM instructions. And this happened at least a few years ago. I don't know where they expected the industry to go then, but I have a good feeling that it didn't play out like they thought it may have. There were some Windows/Android hybrid tablets that were released with zero fanfare, and I think that's a good indication for how much the mainstream markets want ARM/x86 CPUs (and the embedded markets want it considerably less)
Saying that VIA is going to go from selling x86 and ARM SoCs to just x86/ARM hybrid is crazy, their Nano parts are nothing at all like their ARM parts. You can't just take one of these Nano CPUs modified for ARM execution, drop it in a WonderMedia SoC, and call it a day - even if that wasn't a major engineering hurdle for them it'd totally throw it out of the price and power markets those SoCs are designed for.